Article ID: | iaor2000503 |
Country: | Japan |
Volume: | 46 |
Issue: | 2 |
Start Page Number: | 411 |
End Page Number: | 431 |
Publication Date: | Jan 1998 |
Journal: | Proceedings of the Institute of Statistical Mathematics |
Authors: | Shinano Yuji, Fujie Tetsuya |
Keywords: | optimization, programming: mathematical, programming: integer, computational analysis: parallel computers |
The branch-and-bound algorithm is mostly applied for solving combinatorial optimization problems exactly. In practice, however, there is a limit of a problem size which can be solved in a reasonable amount of time. Parallelization of the branch-and-bound algorithm is an important approach to solve more problems beyond the limit, and according to the recent advance of parallel machines, much effort has been devoted to the parallelization. Especially, based on the general framework of the branch-and-bound algorithm, generalized tool development is a trend of the implementations. PUBB (Parallelization Utility for Branch-and-Bound algorithms), developed by Shinano